SAX-VSM: Interpretable Time Series Classification Using SAX and Vector Space Model

摘要

In this paper, we propose a novel method for characteristic patterns discovery in time series. This method, called SAX-VSM, is based on two existing techniques - Symbolic Aggregate approXimation and Vector Space Model. SAX-VSM is capable to automatically discover and rank time series patterns by their importance to the class, which not only creates well-performing classifiers and facilitates clustering, but also provides an interpretable class generalization. The accuracy of the method, as shown through experimental evaluation, is at the level of the current state of the art. While being relatively computationally expensive within a learning phase, our method provides fast, precise, and interpretable classification.